Computing Library › Numerical Methods
Numerical Methods

The Particle-in-Cell Method

A hybrid scheme that follows particles through continuous space while solving fields on a grid, the workhorse of kinetic plasma simulation.

Particles and fields together

The particle-in-cell (PIC) method simulates plasmas at the kinetic level by tracking many computational particles moving through continuous space while computing the electromagnetic fields on a fixed grid. It bridges two descriptions: particles capture the velocity-space physics that a fluid model misses, while the grid makes field solves tractable. Each time step cycles through a fixed sequence of operations coupling the two.

The PIC cycle

Kronos motion — space economy

The particle push

The velocity update must handle the magnetic rotation accurately over many gyrations. The Boris algorithm is the standard pusher: it splits the electric acceleration from the magnetic rotation and applies the rotation exactly, conserving phase-space volume and giving stable long-time orbits. This structure-preserving property is essential when following particles for millions of steps.

Noise and resolution constraints

Because PIC uses a finite number of particles to represent a continuous distribution, it carries statistical noise that decreases only slowly with particle count, similar to Monte Carlo sampling. The grid must resolve the plasma's smallest relevant length (the Debye length) and the time step must resolve the fastest oscillation (the plasma frequency) to avoid numerical heating and instability, a stability requirement analogous to the CFL condition.

Role in fusion

PIC is the primary tool for first-principles kinetic study of plasma instabilities, wave-particle interactions, and non-equilibrium velocity distributions. Its gyrokinetic variant, which averages over the fast gyration, is used to model the turbulence that governs cross-field transport, a key input to predicting confinement in devices like the Hyperion spherical tokamak.